Abstract
Concerning the problems that abnormal detection of network security situation is not easy to be operated with high time and spatial complexity, a quantitative method about hypothesis test has been proposed by acquiring a set of observation samples. Firstly, it can generate network security situation by inference rule from Bayesian network. Then, it can detect whether network security situation is abnormal or not by hypothesis test method, which can forecast the trend of network security situation in future. Thus, it can provide a reliable basis for administrators to make decisions and defensive measures. Finally, taking full use of the simulation data, network abnormal detection algorithm of security situation is verified and the results show that the method is correct and feasible.